Post Thumbnail

Hugging Face challenges DeepSeek: Project Open-R1 reveals secrets of Chinese AI

Hugging Face challenges DeepSeek: Project Open-R1 reveals secrets of Chinese AI The Hugging Face team presented the first results of the Open-R1 project aimed at reproducing the technologies of Chinese artificial intelligence DeepSeek-R1. Within a week, researchers managed to achieve significant progress in understanding and replicating this advanced system.

A key achievement was the successful reproduction of test results on the MATH-500 benchmark. Researchers confirmed impressive performance of various model versions: DeepSeek-R1-Distill-Qwen-32B achieved 95.0% accuracy compared to the claimed 94.3%, while the Llama-70B-based version showed 93.4% versus the official 94.5%.

During the study, a unique feature of DeepSeek-R1 was discovered – unprecedented length of generated responses. Analysis of distribution in the OpenThoughts dataset showed that the average response length is about 6000 tokens, and in some cases exceeds 20,000 tokens. “Considering that an average page contains approximately 500 words, and one token is slightly shorter than a word, many responses exceed 10 pages in volume,” researchers note.

To ensure research transparency, the Hugging Face team created an open Open-R1 leaderboard where the community can track progress in reproducing results. Special attention is paid to the issue of significant GPU memory requirements during training due to the need to generate long sequences.

The Open-R1 project, launched just a week ago, combined the efforts of various teams and the developer community. The main goal remains to reproduce the training pipeline and synthetic data of DeepSeek-R1, which will help better understand the operating principles of this advanced artificial intelligence system.

This initiative demonstrates a growing trend towards openness and collaboration in AI, where even the most complex technological achievements become the subject of collective study and reproduction by the global developer community.

Autor: AIvengo
For 5 years I have been working with machine learning and artificial intelligence. And this field never ceases to amaze, inspire and interest me.

Latest News

Imagry created a drone without HD maps

Imagry company created a unique autonomous control technology that works without HD maps. This engineering marvel is based on a bio-inspired approach that mimics human perception and decision-making. Imagine — the system sees the road with eye-cameras and makes decisions with a brain-neural network, exactly like an experienced driver!

US Department silently replaced report with AI-fake quotes

The US Department of Health and Human Services report on chronic childhood diseases found itself at the center of a scientific scandal. Experts discovered that the document "Make Our Children Healthy Again Assessment" contains falsified quotes and non-existent studies.

Meta AI blocks thousands of Facebook groups for "terrorism"

Earlier I told you that Meta's artificial intelligence was massively blocking Instagram accounts. Now it's blocking Facebook groups too. An unprecedented wave of blocks is happening. Thousands of Facebook groups fell under unfounded sanctions from the moderation system. This technical collapse affected communities both in the USA and abroad, covering the most diverse thematic categories.

Google released Gemini CLI: AI agent for code

Google company presented Gemini CLI. This is an official agent for using artificial intelligence to write code. Directly from the command line. And this tool provides access to all capabilities of the Gemini 2.5 Pro model.

Salesforce test: Gemini 2.5 Pro solves only 58% of business tasks

The Salesforce CRMArena-Pro test shows that even leading artificial intelligence models face serious limitations when solving everyday business tasks.